Lmst

Just coming back from the launch of the European Center for Algorithmic Transparency in Sevilla #ecat #dsa - I wrote some reflections on the future of algorithmic auditing

https://www.get-parity.com/raiblog/algo-auditing-1

@rachelmetz congratulations!!!

@Sbectol@toot.wales absolutely and it’s so bizarre to me!

Rumman Chowdhury boosted:

The artist revolt against AI-powered art generators is in full swing. Artists have filed a class-action lawsuit in California against Stability AI and Midjourney. They've raised money for a D.C. lobbyist. And soon they will have a technological countermeasure, a tool designed by U of Chicago computer scientists that will "cloak" art posted online so that it won't be "feeding this monster that becomes more and more like them." My latest: https://www.nytimes.com/2023/02/13/technology/ai-art-generator-lensa-stable-diffusion.html

Every election cycle I find it fascinating the number of south Asian Americans who think they pass as white because they’re wealthy, did all the “right things” and have white features and hair (if not skin). I’d love to read anything written about it.

Hi all! I’ll be live on @NPR Science Friday tomorrow with Melanie Mitchell of the Santa Fe Institute! We’ll be talking all things AI, including yes, ChatGPT. Tune in and ask us some questions!

@kellogh thanks, this is interesting

When Harvard Law School says you’re greedy, that’s something else https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4330393

Image of abstract from Harvard Law School Paper: Date Written: January 19, 2023

Abstract: This paper provides a case study of the acquisition of Twitter by Elon Musk. Our analysis indicates that when negotiating the sale of their company to Musk, Twitter's leaders chose to disregard the interests of the company's stakeholders and to focus exclusively on the interests of shareholders and the corporate leaders themselves. In particular, Twitter's corporate leaders elected to push under the bus the interests of company employees, as well as the mission statements and core values to which Twitter had pledged allegiance for years. Our analysis supports the view that the stakeholder rhetoric of corporate leaders, including in corporate mission and purpose statements, is mostly for show and is not matched by their actual decisions and conduct (Bebchuk and Tallarita (2020)). Our findings also suggest that corporate leaders selling their company should not be relied upon to safeguard the interests of stakeholders, contrary to the predictions of the implicit promises and team production theories of Coffee (1986), Shleifer- Summers (1988) and Blair-Stout (1999).

@chancerydaily FWIW I prefer you’d talk about it here than there - you drive views and retain users for him. I struggle with it myself.

I’m settling into an equilibrium of using Twitter for big announcements and mastodon for conversation and feedback.

Rumman Chowdhury boosted:

First toot from the official account of Symposium on Usable Privacy and Security! 👋

📢 Call for papers for #SOUPS2023 is out! The paper submission deadline is February 16, 2023. (Make sure to register your submissions by February 10!)
https://www.usenix.org/conference/soups2023/call-for-papers

And please follow us for more updates!

Banner image for publicizing the Nineteenth Symposium on Usable Privacy and Security (SOUPS 2023)

At risk of sounding like a cliche, limiting social media hasn’t left a hole but instead left room for other things to grow. There’s no fomo, no missed connections. I still meet new people and read good papers and learn lots of things. There’s nothing to prove and no one to prove it to.

Rumman Chowdhury boosted:

@mhoye @arclight @otfrom the European Spreadsheet Risk Interest Group has a collection of gloriously horrifying stories starring Excel: https://eusprig.org/research-info/horror-stories/

Their conference is back this year; Chatham House rules and really fun/scary/interesting: https://easychair.org/cfp/EuSpRIG_2023

Rumman Chowdhury boosted:

Decades of research prove that the best way to end homelessness is...

Give people a home. Seriously. That's it. No strings, no prerequisites, no faith-based, holistic, means-tested, yoga-cebtric meditation parole. Just a home.

https://www.newscientist.com/article/2356643-we-can-reduce-homelessness-if-we-follow-the-science-on-what-works/

@coty @bkc Coty!! Good to see you here!

Excited to announce that I’m joining @bkc as. #responsibleai AI fellow! I’ll be working on issues of AI governance.

I’m enthusiastic to be on the outside of industry; there’s a moment now considering Musk and Twitter, tech layoffs, generative AI as well as 2023 being the year we get algorithmic accountability laws (#dsa #EUAIAct)

https://cyber.harvard.edu/story/2023-01/announcing-dr-rumman-chowdhury-berkman-klein-center-responsible-ai-fellow

@kjhealy this should be illegal

Rumman Chowdhury boosted:

New, by me: ODIN Intelligence, the police tech firm whose website was defaced last weekend, was hacked. A huge trove of confidential police data was exfiltrated and provided to transparency collective DDoSecrets. The data contains tactical plans of police raids, and use of surveillance, like facial recognition.

More: https://techcrunch.com/2023/01/21/odin-intelligence-breach-police-surveillance/